Probabilistic Determination of Crash Locations in a Road Network with Imperfect Data

The limited quality of location data poses a major problem to those who want to use such data for research or safety management. The proportion of crashes unassigned to specific road locations is considerable. This fact is sometimes overlooked because crash database queries return crashes assignable to locations but do not issue warnings about crashes that are not assignable. Without reliable location information, crash data are of limited use for more refined safety research and even for basic road safety management. Probabilistic linking techniques are frequently chosen to link medical, census, and other population records, mainly for medical research but also for security concerns and market studies. This paper presents a first application of the probabilistic method of assigning crashes to roads through linking road crash records with road inventory records. Because of missing or incorrect data, multiple locations are candidates, and probabilistic linking offers a solution by selecting the highest-likelihood locations. This paper uses the Bayesian approach to link data in a version specialized for linking one-to-many records.